Classifying and Measuring the Creative Industries

Classifying and Measuring the Creative
Consultation response
Arts Council England
June 2013
Arts Council England welcomes this review of the classification and measurement of the creative
industries and supports the movement towards a more robust and replicable methodology for
classifying and measuring the creative industries.
Along with other members of the Creative Industries Council, we have long been aware of the
inconsistencies and flaws in the existing methodology that this proposal aims to address. The
proposals are welcome, insofar as they aim to address these flaws and increase the robustness of
the existing classifications.
We support the proposed shift to the creative intensities methodology, which we believe has many
strengths. However we share with DCMS the acknowledged concerns: namely, that those sectors
with large support structures are theoretically more likely to be excluded from the estimates, and
that the underlying SIC/SOC framework on which the methodology rests is not adequately granular
to capture the range of activity with the creative industries.
For the Arts Council, this presents four key challenges, which can be broadly grouped under the
following headings:
the crafts sector
the retail sale of art
Below, we explore the particular issues relating to these four sectors, and where possible propose
some ways in which the methodology (or its application) might be adapted.
We welcome DCMS’s transparency and openness during this process, and believe that this
consultation has been positive for our sector, not least in increasing understanding and engagement
in the question of measurement.
We acknowledge that there has been some degree of unease about the proposals, and in particular
the risk of exclusion of some sub-sectors from the department’s creative industries definitions. Due
to the economic context, and the emphasis by Government on promoting growth, all creative
industry sectors are operating in uncertain times, and we know that the ability to accurately
measure economic indicators is linked, more than ever, to sustainability and success. In that
context, we welcome the Department’s recent clarification in regards to crafts, and look forward to
proactively working together over the long-term in pursuit of better and more inclusive ways to
measure the creative industries.
Consultation issue 1: what are your views of using the Creative Intensities methodology to underpin
the DCMS Creative Industries classification?
Please note that our specific concerns with the outcomes of the methodology’s application are
addressed under consultation questions two, three and four below. This section concerns the
methodology itself.
We support the use of the creative intensities methodology to underpin the DCMS Creative
Industries classifications.
We believe that the time is right to improve the existing classifications. We are aware that there is a
longer term initiative to propose amendments to the SIC/SOC framework, in order that it might
better reflect activity within the creative industries. Given that any change to the SIC/SOC
framework is certain to be many years away, we would not support unnecessary delay in the
improvement of the Creative Industries estimates on this basis.
Below, we comment upon each stage of the methodology in turn.
1. Identifying creative occupations
We support the reversal of the existing methodology, in that it begins by identifying creative
occupations. This method is bottom-up, and does not suffer from the flaws in the current
methodology, which includes a number of areas in which labour is non-creative.
We support the exclusion of SOC codes where the vast majority of labour within that code is noncreative. We acknowledge that due to the imperfect granularity of SOC codes, this necessarily
means the exclusion of some creative occupations.
Our main concern with this stage of the methodology is the extent to which it is sustainable in the
long-term. The list of occupations outlined in consultation document is based upon the professional
judgment of Creative Skillset and others, in their paper Classifying and measuring the creative
industries. Whilst the proposed list is broadly accurate, in order for it to remain up-to-date it is
certain to need regular review, in the context of a changing workforce (especially in digital-related
activities) and a potential shake-up of the SOC framework itself. The current methodology does not
make it clear how this will happen, who will do it, when it might happen, how regularly, nor what
criteria will be used.
With this in mind, we would support the development of a plan for the future review of this list, and
the adoption of a set of criteria, similar to those outlined by Nesta in their paper A dynamic mapping
of the UK’s creative industries. Although the two lists of occupations produced by Creative Skillset
and Nesta are nearly identical, we favour the more transparent approach adopted by Nesta in
developing their list.
We believe Nesta’s five criteria are broadly sound, but would ask to be involved in a review of these,
were they to be formally adopted going forward.
2. Measuring employment in these creative occupations in each sector to identify those that can
be classified as creative industries
We support the use of employment as an identifying measure of the creative industries, as part of
this methodology, not least because it puts creative talent and skills at the heart of the creative
We support the decision not to move towards a different set of identifiers, such as intellectual
property rights generation.
We support the decision to move from proportional system to an ‘in or out’ approach for SIC
sectors. We note that in doing so, this will necessarily exclude many who would otherwise be ‘in’,
and (to a lesser extent) include those who would otherwise be ‘out’. Nevertheless, we recognise
that the proportional system was deeply flawed, and was not consistent with the aim of producing a
robust, replicable methodology, which avoids double-counting at a national level.
We cautiously support the use of the 30% threshold at which a sector is deemed to be creative. We
recognise this as problematic, for the same reason that employment data is theoretically an
imperfect tool, in that there are clearly industries in which the creative employment is below 30%,
but where the central value produced by the industry is clearly ‘creative’. Museums are an excellent
example of this (we discuss this issue later).
However, the problem here lies in the adoption of a threshold itself, not in the number chosen to be
that threshold (30%). On the basis that a threshold exists, which it clearly should, we believe 30% to
be a prudent level at which to set it. We believe that the method which produced this number
grants it credibility, and we are reliant on the distributional analysis undertaken by Nesta.
3. Grouping these SIC sectors into broad creative industries groups
We support the grouping of SIC codes into broad creative industries groups, and note that, with a
few exceptions, the proposed new list of SIC sectors maps well onto the broad groups that we
traditionally think of as making up the creative industries.
We note that whilst the sector groupings will be useful for the reporting of economic indicators, they
should not be confused with DCMS’s definition of, or list of interests within, the creative industries.
DCMS will continue to have policy responsibility for those sectors removed from these sector
groupings (e.g. ‘Crafts’ and ‘Arts and Antiques’) and in the interests of avoiding confusion, should
find a way to indicate this. In publication and presentation, we would thus support a more specific
and unique titling than ‘broad creative industry groups’.
Consultation issue 2: what are your views of the list of Creative Occupations as defined above? Are
there occupations which have been included which you think should not be? Are there occupations
which have not been included which you think should be? What evidence do you have (if any) to
support your view on inclusions or exclusions?
Consultation issue 3: what are your views of the list of sectors as defined above? Are there sectors
which have been included which you think should not be? Are there sectors which have not been
included which you think should be? What evidence do you have (if any) to support your view on
inclusions or exclusions?
Note: we have combined our responses to issues 2 and 3, as it was not possible to consider these
issues in isolation.
Our specific concerns about the proposed list of occupations and industry sectors are fourfold, and
all relate to industries that are excluded from the proposed list in Figure 8. At a broad sector level,
these can be summarised as:
Retail sale of art
We take each of these in turn below.
We acknowledge that the difficulties in measuring crafts are highlighted in the consultation
document, and have been the subject of some debate and controversy in the last two months.
Setting aside the technical issue of capturing indicators for a sector which is formed primarily of selfemployed workers and micro businesses, the main difficulty in measuring crafts exists at SOC level.
This is because the SOC codes delineate craft-based activity by type of product/output, not by (for
instance) scale of output, ‘cultural value’ of output, type of labour, skills, etc., or other ways in which
we commonly distinguish between manufacturing and making. To put it another way, SOC codes
allow us to broadly distinguish between those that work with glass and those that work with wood,
but does not allow us to distinguish between those that work creatively with wood, and those that
do not.
In the updated SOC2010 list, the main SOC codes for craft activity are within the two-digit code 54:
Textiles, Printing and Other Skilled Trades, as well as additional four-digit codes from elsewhere
within the single-digit 5 code (Skills trades occupations).
Currently, none of these codes are included in the DCMS proposals, and we understand the essential
reasoning behind this. However we believe that upon closer inspection several of these SOC codes
are less imperfect than has been previously assumed, and that there is a case for their inclusion,
given that:
a) The skills required for the majority of the activity under these codes is correlative to those
required by an artist working in that material, and that in application of those skills, the worker is
likely to be engaging in creative labour, and
b) Many of these codes have an industrial- or construction-type equivalent, held within the codes
under 81 (Process operatives) and 53 (Skilled construction and building trades) which allows us to
more easily distinguish hands-on interpretive activity.
For instance, it is easier to separate out industrial- or construction-type manufacturing within glass
and ceramic production, as it is otherwise captured under code 8112 (Glass And Ceramics Process
Operatives), or 5316 (Glaziers, window fabricators and fitters). As such, the remaining employment
within 5441 is likely to retain a hands-on, interpretive nature, which in our view is creative. We also
note that in Nesta’s analysis, this code scores the maximum (5) when assessed against their criteria
and is included on that basis.
In this we have also been guided by research commissioned by the Crafts Council1 who note that the
aggregation of crafts occupations that occurred during the update from SOC2000 to SOC2010 has
had less of an effect on the crafts coding as has been previously assumed, i.e., the catch-all code
5449 (Other Skilled Trades n.e.c.) is found to contain a high number of occupations previously
identified as creative.
As such, we believe that the list of creative SOC codes could be broadened to include:
Smiths and forge workers
Weavers and knitters
Print finishing and binding workers
Glass and ceramics makers, decorators and finishers
Furniture makers and other craft woodworkers
Other skilled trades n.e.c.
Of course, we are aware that the inclusion of further SOC codes is not guaranteed to impact upon
the inclusion of further SIC codes to the sector list. Nesta’s analysis suggests the possible inclusion of
SIC codes 3212 (Manufacture of jewellery and related articles) and 2341 (Manufacture of ceramic
household and ornamental articles) on the basis of high creative intensity. However, TBR’s analysis
suggests that only one code, SIC 3109 (Manufacture of other furniture), would be included, once the
minimum employment size threshold (10,000) is applied.
In this, we have two concerns:
Firstly, that if a single craft SIC code (3109) is included in the list of creative industries, it presents
difficulties for the eventual grouping of these codes into broad sector categories. Either ‘Craft’
would be represented be only one (relatively small) SIC code, which forms a small strand of the total
TBR, Defining and Measuring Craft, 2013. For space, we have avoided repeating the analysis done by TBR
here. We advise reading this paper for a lucid exploration of this issue.
activity within it; or else the code is included in another grouping, in which it will certainly be out of
place. In our opinion, neither of these is ideal; but the first is preferable.
Secondly, were this to be corrected by amending the minimum employment size threshold by which
industries could be included, we would be highly concerned about the reliability of the data and the
usefulness of the eventual estimates. We remain unconvinced by the plausibility of reducing the
minimum employment size threshold, but believe this could be looked at again were changes
implemented by ONS to make it more feasible.
We believe that the route towards a truly robust measurement of indicators for the crafts sector will
lie with a correction of the underlying SIC/SOC framework and improvements to data collection
methods. We will work proactively with DCMS, the crafts sector and the Crafts Council in pursuit of
Retail sale of art
We note that the new methodology will exclude art dealing from its definitions; specifically, the
removal of the SIC code 47.78/1 (Retail sale in commercial art galleries), on the basis of low intensity
of creative workers. The underlying issue here is similar to that suffered by the craft sector, in that
the flaw is contained within the SOC structure itself:
The majority of art dealers are likely to be contained within codes that refer primarily to retail
activity, and which thus do not capture the creativity of their labour, nor the creative value chain in
which they are a part. Art dealers will primarily be classified either by the four-digit 1254
(Shopkeepers and proprietors – wholesale and retail) or 1190 (Managers and directors in retail and
wholesale), or otherwise in one of the codes under the two-digit 71 (Sales occupations). These
codes do not make adequate distinction between the type, scale or value of sale.
Arts Council England believes that at a policy level, the retail sale of art is inseparably entwined with
artistic creation, and success in one will be linked to success in the other. Nevertheless, we support
the principle of excluding occupations in which the majority of activity is non-creative. As such, we
support the removal of the SIC code 47.78/1 (Retail sale in commercial art galleries) from the list of
creative industries.
Museums and libraries
Note: Although these sectors are very different, we combine them here as the issues encountered
with the methodology are the same.
We note that museums and libraries are excluded from the list of creative industries (Figure 2).
Specifically that SIC 91.01 (Library and archive activities) and 91.02 (Museum activities) are excluded
on the basis of low creative intensity.
Although there are a variety of issues with the accurate measurement of indicators related to
museums,2 particularly related to the way that data is collected, the problem encountered here, is
related to the proposed methodology itself.
Whilst the DCMS’s assessment of occupation codes 2451 (Librarians) and 2452 (Archivists and
curators) finds them to be creative (with which we concur), the percentage of these occupations
within the industry codes SIC 91.01 (Library and archive activities) and 91.02 (Museum activities) is
not sufficiently high enough for the industries to cross the 30% threshold, due to the high number of
support staff that these two industries employ.
We strongly support a classification of the creative industries which includes these two industries.
There are two ways in which one might theoretically address this:
Firstly, within the methodology, one might revise the threshold at which industries are deemed to be
creative. For museums and libraries, a revision from 30% to 20% would address their exclusion.
We do not support this proposal for two reasons: firstly, any movement of the threshold affects not
just those industries we have identified, but also all others with a similar intensity level. A shift as
significant as 10%, as above, would have profound effects on the list of industries included in the
classifications. The problem of incorrect inclusion and exclusion is a structural problem of the
threshold itself, not the percentage at which it is fixed. Nesta have undertaken in-depth
distributional analysis of different threshold levels, and we believe 30% to be a robust number to fix
a threshold. The second reason is that an arbitrary shift in the threshold does not guarantee the
continuing inclusion of museums and libraries into the creative industries. Between 2009 and 2010,
the creative intensity for SIC 91.01 (Library and archive activities) reduced from 27% to 23%. There
is no guarantee that were the threshold shifted to 20%, the sector intensities would stay above that
The second way that one might address the exclusion of SIC 91.01 (Library and archive activities) and
91.02 (Museum activities), after the application of the methodology, is to perform a qualitative
secondary stage in the identification of creative industry sectors.
Assuming that the creative intensity methodology is a baseline, upon which we might further amend
the classifications (which the Creative Skillset report would appear to indicate), we believe this to be
the soundest approach. Following the identification of creative industry sectors using the creative
intensities model, a further non-data-based stage could be included wherein a small number of
industry sectors, identified through professional judgment, were included.
This stage would need to be performed against sound criteria, and could only include industries
which were excluded on the basis of the 30% threshold alone. Theoretically, this could also include a
further bottom threshold, by which a sector was deemed beyond this stage (for instance, above
In a recent study commissioned by Arts Council England and NMDC on the economic contribution of the arts
and culture (full details under section 6) we uncovered additional challenges in accurately measuring
indicators related to museums. The research found that the majority of museums (1500 of 2000) were not
captured under SIC 91.02 (museums activities) by the Annual Business Survey, due largely to the fact that
museums often classify their subsidiary trading companies under different codes to their trust or foundation
companies. A full explanation of the issue can be found here (pp.10-11). Arts Council England are working
with NMDC to explore possible solutions to this in the long-term.
15%). We recognise that as a non-data based exercise, this proposal is not wholly commensurate
with the move towards a robust and replicable method of classification; however we believe there is
a balance to be struck on this issue.
Arts Council England believes it right to include to include the two industry codes SIC 91.01 (Library
and archive activities) and 91.02 (Museum activities), on this basis.
Additional notes:
The Creative Skillset report notes that there is possible challenge to the inclusion of the SIC code
74.30 (Translation and interpretation activities) on the basis that the SOC code 34.12 (Authors,
writers and translators) does not adequately disaggregate authors from translators.
Whilst we recognise that there are likely to be many working in the field of ‘translation’ (in its
broadest sense) that are not doing so with a creative remit, we believe that translation work is often
highly creative, particularly (but not only) in the field of literature. We support the inclusion of SIC
code 74.30 (Translation and interpretation activities).
IT, software and computer services:
We welcome the inclusion of additional SIC codes related to IT, software and computer services, in
recognition of their creativity, and the strength of the classifications in this area.
Cultural education:
We strongly support the proposed inclusion of the SIC code 85.52 Cultural Education.
Consultation issue 4: what are your views of the list of groupings as shown above? Are there other
groupings which you think would be preferable, given the SIC codes available to match together in
these groupings?
We believe that the groupings outlined in the consultation document are adequate, given the SIC
codes available, and we have no suggestions for improvement.
We note again that whilst the sector groupings will be useful for the reporting of economic
indicators, they should not be confused with DCMS’s definition of, or list of interests within, the
creative industries. DCMS will continue to have policy responsibility for those sectors removed from
these sector groupings (e.g. ‘Crafts’ and ‘Arts and Antiques’) and should continue to reflect this in its
stated remit. In publication and presentation, we would thus support a more specific and unique
titling than ‘broad creative industry groups’.
Consultation issue 5: do the SIC and SOC codes adequately and accurately capture the full range of
economic activity within the creative industries. If not, how would you better define the SIC and SOC
Clearly, the SIC and SOC codes are far from perfect, and there are well-known issues with their
ability to adequately describe the activity within the creative industries, both in terms of their
comprehensiveness and granularity.
The list of challenges are too numerous to detail here, and we have decided to limit our response
here to issues produced directly as a result of DCMS’s creative industries proposed methodology.
We will continue to work proactively with DCMS on identifying improvements to the industrial and
occupational classification systems. We are also aware of a future consultation on this issue, led by
Creative Skillset, and Arts Council England will submit a response in that forum.
Consultation issue 6: What indicators do you find to be of value in your work? Do you collect data
against these indicators on a regular basis and, if so, how do you do so?
We find the current set of indicators used by DCMS highly valuable, and have no suggestions for
further inclusions at this stage.
Until recently, Arts Council England has relied almost solely upon figures supplied by DCMS’s
economic estimates, appended by data from CC Skills.
Earlier this year, Arts Council England commissioned the Centre for Economics and Business
Research to explore the contribution of arts and culture to the national economy. Their report, The
contribution of the arts and culture to the national economy, can be found here.
Their research found that arts and culture generated £12.4 billion turnover in 2011, including just
under £6 billion of Gross Value Added (GVA) to the UK economy. It also demonstrated that the
sector accounts for 110,600 FTE jobs in the UK in 2008-2011 (before multiplier impacts are added).
The research also found that 90 – 94% of businesses operating within the arts and culture industry
are made up of firms with one or no employees. This would suggest that there may be a high
proportion of businesses within our sector which are unlikely to be captured by official datasets,
emphasising the need for a holistic approach to improving economic measurements for the arts and
cultural sectors.
If you have any questions regarding this submission, please contact:
Ben Brown
Officer, Policy Development
+44 (0)20 7973 6413